Accurate long-range forecasting of COVID-19 mortality in the USA

Abstract The need for improved models that can accurately predict COVID-19 dynamics is vital to managing the pandemic and its consequences. We use machine learning techniques to design an adaptive learner that, based on epidemiological data available at any given time, produces a model that accurate...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Pouria Ramazi, Arezoo Haratian, Maryam Meghdadi, Arash Mari Oriyad, Mark A. Lewis, Zeinab Maleki, Roberto Vega, Hao Wang, David S. Wishart, Russell Greiner
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/ac72519b371340d4aede2c0064a6cd98
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ac72519b371340d4aede2c0064a6cd98
record_format	dspace
spelling	oai:doaj.org-article:ac72519b371340d4aede2c0064a6cd982021-12-02T15:23:07ZAccurate long-range forecasting of COVID-19 mortality in the USA10.1038/s41598-021-91365-22045-2322https://doaj.org/article/ac72519b371340d4aede2c0064a6cd982021-07-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-91365-2https://doaj.org/toc/2045-2322Abstract The need for improved models that can accurately predict COVID-19 dynamics is vital to managing the pandemic and its consequences. We use machine learning techniques to design an adaptive learner that, based on epidemiological data available at any given time, produces a model that accurately forecasts the number of reported COVID-19 deaths and cases in the United States, up to 10 weeks into the future with a mean absolute percentage error of 9%. In addition to being the most accurate long-range COVID predictor so far developed, it captures the observed periodicity in daily reported numbers. Its effectiveness is based on three design features: (1) producing different model parameters to predict the number of COVID deaths (and cases) from each time and for a given number of weeks into the future, (2) systematically searching over the available covariates and their historical values to find an effective combination, and (3) training the model using “last-fold partitioning”, where each proposed model is validated on only the last instance of the training dataset, rather than being cross-validated. Assessments against many other published COVID predictors show that this predictor is 19–48% more accurate.Pouria RamaziArezoo HaratianMaryam MeghdadiArash Mari OriyadMark A. LewisZeinab MalekiRoberto VegaHao WangDavid S. WishartRussell GreinerNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Pouria Ramazi Arezoo Haratian Maryam Meghdadi Arash Mari Oriyad Mark A. Lewis Zeinab Maleki Roberto Vega Hao Wang David S. Wishart Russell Greiner Accurate long-range forecasting of COVID-19 mortality in the USA
description	Abstract The need for improved models that can accurately predict COVID-19 dynamics is vital to managing the pandemic and its consequences. We use machine learning techniques to design an adaptive learner that, based on epidemiological data available at any given time, produces a model that accurately forecasts the number of reported COVID-19 deaths and cases in the United States, up to 10 weeks into the future with a mean absolute percentage error of 9%. In addition to being the most accurate long-range COVID predictor so far developed, it captures the observed periodicity in daily reported numbers. Its effectiveness is based on three design features: (1) producing different model parameters to predict the number of COVID deaths (and cases) from each time and for a given number of weeks into the future, (2) systematically searching over the available covariates and their historical values to find an effective combination, and (3) training the model using “last-fold partitioning”, where each proposed model is validated on only the last instance of the training dataset, rather than being cross-validated. Assessments against many other published COVID predictors show that this predictor is 19–48% more accurate.
format	article
author	Pouria Ramazi Arezoo Haratian Maryam Meghdadi Arash Mari Oriyad Mark A. Lewis Zeinab Maleki Roberto Vega Hao Wang David S. Wishart Russell Greiner
author_facet	Pouria Ramazi Arezoo Haratian Maryam Meghdadi Arash Mari Oriyad Mark A. Lewis Zeinab Maleki Roberto Vega Hao Wang David S. Wishart Russell Greiner
author_sort	Pouria Ramazi
title	Accurate long-range forecasting of COVID-19 mortality in the USA
title_short	Accurate long-range forecasting of COVID-19 mortality in the USA
title_full	Accurate long-range forecasting of COVID-19 mortality in the USA
title_fullStr	Accurate long-range forecasting of COVID-19 mortality in the USA
title_full_unstemmed	Accurate long-range forecasting of COVID-19 mortality in the USA
title_sort	accurate long-range forecasting of covid-19 mortality in the usa
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/ac72519b371340d4aede2c0064a6cd98
work_keys_str_mv	AT pouriaramazi accuratelongrangeforecastingofcovid19mortalityintheusa AT arezooharatian accuratelongrangeforecastingofcovid19mortalityintheusa AT maryammeghdadi accuratelongrangeforecastingofcovid19mortalityintheusa AT arashmarioriyad accuratelongrangeforecastingofcovid19mortalityintheusa AT markalewis accuratelongrangeforecastingofcovid19mortalityintheusa AT zeinabmaleki accuratelongrangeforecastingofcovid19mortalityintheusa AT robertovega accuratelongrangeforecastingofcovid19mortalityintheusa AT haowang accuratelongrangeforecastingofcovid19mortalityintheusa AT davidswishart accuratelongrangeforecastingofcovid19mortalityintheusa AT russellgreiner accuratelongrangeforecastingofcovid19mortalityintheusa
_version_	1718387347351928832

Accurate long-range forecasting of COVID-19 mortality in the USA

Ejemplares similares